Epstein-Barr virus (EBV) is a human herpesvirus that causes infectious mononucleosis. Prospective epidemiologic studies indicate that EBV is causally associated with at least two malignancies - endemic Burkitt's lymphoma, and nasopharyngeal carcinoma. EBV's genome is latent in all these diseases, and is maintained as a nuclear episome that replicates once per cell cycle, and is efficiently partitioned at mitosis. Understanding the mechanism by which EBV episomes are replicated once per cell-cycle and partitioned will ultimately yield therapies against all diseases caused by a latent EBV infection. The long-range goal of these studies is to understand how viral episomes in general, but EBV genomes in specific, are maintained in a functional state with the capacity to cause human disease. The only viral protein that is required for EBV genome maintenance and partitioning is the Epstein-Barr nuclear antigen 1 (EBNA1), which binds a viral cis-element termed oriP through its carboxy-terminal DNA-binding domain. OriP contains two sets of binding sites for EBNA1 - DS, a region that serves as an origin of replication, and FR, a region that serves as an episome maintenance and partitioning element when EBNA1 is bound to it. We have recently discovered that the amino-terminal domain of EBNA1 has two regions within it that are AT-hooks, corresponding to the AT-hooks of cellular proteins such as the prototypic AT-hook protein HMGA1a. Confirming this observation, we have demonstrated that we can functionally replace the N-terminus of EBNA1 with the AT-hooks of HMGA1a. In this proposal we propose to extend our studies in three areas: 1. To characterize the roles of the AT-hook regions of EBNA1 in chromosome tethering and episome maintenance. 2. To utilize a novel replicon and fusion protein to understand the function of EBNA1 in replication. 3. To use viral genetics to determine the functions of EBNA1 in the immortalization of naive B-cells by EBV